[Gandalf Parker]

Interestingly even before the nerf, on massive AI tests it was not the heims that came out on top. I dont even remember them winning any of the games tho they probably did.

The tests were like endurance tests for the game. All of the nations were in as AI's, the game was run on immeadiate hosting and went for days. Mostly I watched for crashes (of course), CPU loads, hosting time (it started at about 1 hosting every 10 minutes and went to alittle over 30 minutes), and memory leaks. Each game did create a log and a scores page. I did not keep great notes about the winners but I was looking to see if the same nation came up too often. There was nothing worrisome to report to the beta team.

Different maps, map sizes, victory conditions, other settings did make a difference in the game. In abit above average times on standard settings; the games tended to be Undead Ermor doing well in the early game, Oceania in mid game, and Arcos in late game if it had managed to get there.

I had considered automating it all and doing a big batch of one-on-one contests but I never got motivated enough to do it.